Characterizing effects of hydropower plants on sub-daily flow regimes

María Dolores Bejarano , Álvaro Sordo-Ward , Carlos Alonso , Christer Nilsson

a b s t r a c t

A characterization of short-term changes in river flow is essential for understanding the ecological effects of hydropower plants,
which operate by turning the turbines on or off to generate electricity following variations in the market demand (i.e.,
hydropeaking). The goal of our study was to develop an approach for characterizing the effects of hydropower plant operations on
within-day flow regimes across multiple dams and rivers. For this aim we first defined ecologically meaningful metrics that
provide a full repre-sentation of the flow regime at short time scales from free-flowing rivers and rivers exposed to hydropeaking.
We then defined metrics that enable quantification of the deviation of the altered short-term flow regime variables from those of
the unaltered state. The approach was successfully tested in two rivers in northern Sweden, one free-flowing and another
regulated by cascades of hydropower plants, which were additionally classified based on their impact on short-term flows in sites
of similar management. The largest differences between study sites corresponded to metrics describing sub-daily flow
magnitudes such as amplitude (i.e., difference between the highest and the lowest hourly flows) and rates (i.e., rise and fall rates
of hourly flows). They were closely followed by frequency-related met-rics accounting for the numbers of within-day hourly flow
patterns (i.e., rises, falls and periods of stability of hourly flows). In comparison, between-site differences for the duration-related
metrics were smallest. In general, hydropeaking resulted in higher within-day flow amplitudes and rates and more but shorter
periods of a similar hourly flow patterns per day. The impacted flow feature and the characteristics of the impact (i.e., intensity
and whether the impact increases or decreases whatever is being described by the metric) varied with season. Our approach is
useful for catchment management planning, defining environmental flow targets, prioritizing river restoration or dam
reoperation efforts and contributing information for relicensing hydropower dams.

Keywords:
Hydrological alterations
Hydrological characterization
Hydropeaking
Impact assessment
Short-term
Sub-daily flows

1. Introduction variation in flows associated to seasons (Bejarano et al., 2010)

are examples of large time-scale flow variability. Additionally,
Critical components of the flow regime such as topography and geology are usually superimposed on climate
magnitude, fre-quency, duration, timing and rate of and shape intra-annual flow variation in, for example, snowmelt-
change control ecological pro-cesses in river fed or groundwater-fed rivers (Bejarano et al., 2010). Furthermore,
ecosystems (Poff et al., 1997), and modification of flow flow variability may also be considered at shorter time scales, from
regimes constrains the distribution of species, their months to hours (or smaller). Day-to-day and within-day water
adaptive capacity, survival, dispersal and reproduction gains or losses are ultimately caused by varying rates of precipita-
(Lytle and Poff, 2004). Each of these five flow tion, evapotranspiration, infiltration, and snowmelt and by catch-
components describes the variability over a wide range ment characteristics such as drainage area, slope and land uses
of spatial and temporal scales (Ward, 1989). Flow (Lundquist and Cayan, 2002; Archer and Newson, 2002), and can
variability may be considered at long time scales, often be in the order of 10% of the mean daily flow in free-
which are commonly controlled by inter- and intra- flowing rivers (Schuster et al., 2008). While these variations are
annual variations in cli-mate. Year-to-year variation in small relative to the variability at annual time scales, they are still
flows associated to the Interdecadal Pacific Oscillation likely to be important to some stream ecosystem characteristics.
index and shifts in the El Niño Southern Oscilla-tion Biggs et al. (2005) described how flow variation at these different
phenomenon (Biggs et al., 2005), and month-to-month temporal scales affects different ecosystem components and
processes in rivers from New Zealand. They recognized that there characterization of flow regimes at such shorter time scales and
may be a hierarchical relationship between time scales of flow for evaluating the intensity of the changes.
variability and related physical processes, the effect of these phys- The fact that these pairs of flow series are commonly difficult to
ical processes on biological processes and, ultimately, the organiza- find might have discouraged the studies on short-term flow
tion of ecosystem characteristics. regimes up to date, though this situation is reverting in recent
Rivers used for hydropower production usually show day-to- times. Thus, new methods are needed to comprehensively describe
day and within-day flow variations that are considerably higher, all facets of within-day flow regimes and assess their degree of
more rapid and frequent than the ones characterizing free- deviation from the natural conditions, to identify dams that artifi-
flowing rivers. This is the result of turning hydro-turbines on or cially modify natural sub-daily variations and river reaches that are
off to generate electricity based on variations in the market likely to experience ecological degradation because of it. Such anal-
demand, so called hydropeaking (Moog, 1993), which has been yses are useful for catchment management plans, defining environ-
recently promoted by the deregulation of the energy market. Addi- mental flow targets, prioritizing river restoration or dam
tionally, changes in the short-term flow regimes are accompanied reoperation efforts and contributing information for relicensing
by changes in hydraulic parameters such as water level, flow veloc- hydropower dams. The goal of our study was to develop an
ity and bed shear stress, and in water quality and river morphol- approach for assessing the effects of hydropower dam operations
ogy, and all together cause significant environmental losses in on within-day flow regimes across multiple dams and rivers. For
the fluvial systems. Although there are still many unknowns, stud- this aim we first defined ecologically meaningful metrics that pro-
ies have revealed significant effects of hydropeaking on fish, vide a full representation of the short-term variation of flow in
including low egg survival (Casas-Mulet et al., 2015), slow growth free-flowing rivers and rivers exposed to hydropeaking. We then
(Flodmark et al., 2004), reduced abundance (Korman and Campana, defined metrics that enable quantification of the deviation of the
2009), stranding (Saltveit et al., 2001), habitat deterioration characterized altered short-term flow regime from the unaltered
(Vehanen et al., 2005) and changes in behavior (Robertson et al., state. We applied devised characterization and impact metrics to
2004). A few studies have also pointed out heavy drift of macroin- several study sites along a free-flowing river and a river with
vertebrates (Carolli et al., 2012), and reductions in the occurrences hydropeaking (at hydropower plant locations) and, with manage-
of beetles (Van Looy et al., 2007) and macrophytes (Mjelde et al., ment facilitation purposes, we classified them according to their
2013). Above all, hydropower is the world’s leading form of renew- short-term flow regime alterations.
able energy, and its demand is likely to increase globally as being a
clean, flexible, and renewable energy source which does not pro- 2. Material and methods
duce greenhouse gases. Development of new hydropower plants
is accelerating in Southeast Asia, Africa, and Latin America (Jager 2.1. Study area and flow data
et al., 2015). In Europe, hydropower is being promoted by legisla-
tion such as the Renewable Energy Directive (RES; 2009/28/EC), The study was located to the Vindel and Ume rivers in the Ume
which sets a legally binding national target of 20% of gross final River basin in northern Sweden (Fig. 1). The Vindel River is the
energy consumption from renewable sources by 2020. In addition, main tributary of the Ume River; it runs parallel to the Ume and
in northern countries, climate change models predict future hydro- joins it about 30 km upstream of the mouth in the Baltic Sea. Both
graphs to match power demands better, increasing the potential rivers show similar characteristics. The whole Ume basin is charac-
for producing more electricity (European Greenpower Marketing, terized by cold-temperate climate, boreal coniferous vegetation
2006). Consequently, an important challenge for river manage- and podzol soils. The upland vegetation consists of subalpine birch
ment arises which involves maximizing hydropower production forests dominated by Betula pubescens, and coniferous forests dom-
with minor ecological impacts. To cope with this demand for inated by Pinus sylvestris and Picea abies. The riparian vegetation
industry and society, assessment of the short-term changes in river includes woody species such as Alnus incana, B. pubescens and Salix
flow following hydropeaking and of the resulting ecological spp., and herbs such as Carex spp. and Ranunculus reptans. The Vin-
responses is key. This paper deals with such assessment. del and Ume rivers have catchment areas encompassing 13,183
To evaluate the impact of hydropeaking resulting from hydro- and 13,633 km2, respectively, their channel lengths are 445 and
power production on short-term (e.g., sub-daily) flow regimes, it 455 km, and their natural mean monthly flows (at the junction)
is necessary to characterize the within-day flow regime along the 197 and 239 m3/s. Whereas the flow regime of the Vindel River
river reach affected by the hydropower plant and to quantify its remains unaltered, the Ume River flow is highly impacted by a
deviation from the unaltered state. Metrics available are scarce chain of hydropower plants and reservoirs which cause hydropeak-
and do not allow a comprehensive characterization of short-term ing (Fig. 2). The free-flowing regime experiences a marked seasonal
flow regimes as they do not account for all hydrological attributes variation with low flows during late autumn and winter and floods
of ecological importance (Zimmerman et al., 2010; Meile et al., during spring. Within a day, the free-flowing regime is relatively
2011; Haas et al., 2014; Sauterleute and Charmasson, 2014; smooth and only fluctuates significantly after water additions or
Bevelhimer et al., 2015; Carolli et al., 2015; Chen et al., 2015). In losses resulting from significant precipitation, evapotranspiration,
addition, most proposed metrics are not conceived to quantify infiltration and snowmelt events. In contrast, dams and reservoirs
the degree of alteration. Research to date has focused on flow vari- alter both the long- and short-term flow regimes of the Ume River;
ability at the daily, seasonal and longer time scales (see review by whereas the natural seasonality of flows is attenuated, the within-
Olden and Poff, 2003). Most characterizations of flow regimes, day flows fluctuate abruptly (Fig. 2). We selected three sites along
quantitative measures of their alterations, and tools and software the Vindel River [from upstream to downstream: Gautsträsk (U;
available for calculations are based on daily-averaged flow records 33 m3/s mean annual flow), Sorsele (S; 119 m3/s) and Granåker
(e.g. Richter et al., 1996, 1997; Clausen and Biggs, 2000; Baker (K; 176 m3/s)] and eight sites along the Ume River coinciding with
et al., 2004; Gao et al., 2009; Carlisle et al., 2011; Fitzhugh and dam and reservoir locations [Grundfors (G; 187 m3/s), Rusfors (R;
Vogel, 2011), which are not precise enough to capture key 213 m3/s), Bålforsen (L; 215 m3/s), Betsele (B; 218 m3/s), Tuggen
components of sub-daily flow fluctuation. Long series of instanta- (T; 222 m3/s), Bjurfors övre (O; 227 m3/s), Bjurfors nedre (N;
neous flow records (e.g., every 15, 30 or 60 min) are required from 232 m3/s) and Harrsele (H; 235 m3/s)] where 15-min and 1-h
both the altered and comparable free-flowing conditions for interval flows were available, respectively (Fig. 1). For the
 Fig. 1. Location map. The name of the sites where the proposed approach was tested (grey circles) are indicated on the map.

short-term flow regime characterization and alteration assessment intervals, named Short-Term Characterization Metrics (STCM;
we chose a 9-year period (2003–2011) of sub-daily flow records, Table 1). For the metrics definition, the within-day hydrograph
common to both rivers. was divided into hours (H) and periods (P). A within-day hourly
hydrograph is characterized by 24 h (H) with individual flow
2.2. Short-term flow regime characterization and alteration records, Qh, which can be assigned one of the following patterns:
assessment (1) rise (RH), when Q(h)
Q(h-1) > 0; (2) fall (FH),when Q(h)
Q(h-1) < 0;
(3) stability (SH) when Q(h)
Q(h-1) = 0; (4) change (CH), when
Our first goal was to characterize the daily flow variation. We the pattern in Q(h
1)– the pattern in Q(h+1); (5) minimum (MinH),
therefore defined a series of hydrological metrics based on 1-h flow when Q(h) = Q(min); and (6) maximum (MaxH), when Q(h) = Q(max).
 a) Ume River (Harrsele)
1600 Vindel River (Granaker)
1400

Flow (m/s)
1200

3
1000
800
600
400
200
0
Ja Fe Ma Ap Ma Ju Jl Au Se Oc No De
b)
500
450
400
350
Flow (m/s)
3

300
250
200
150
100
50
0
Mo Tu We Th Fr Sa Su
Fig. 2. Example of (a) the long-term flow regime and (b) short-term flow regime for the free-flowing Vindel River (black) and the Ume River (grey), which is used for
hydropower production. The x-axis shows (a) the months and (b) the days of the week.

Where h is the 1-h time step and min, and max are the daily hourly We aimed for an evaluation of the short-term flow-regime
minimum and maximum, respectively. One to several periods (P) alteration of the river reaches subjected to hydropower production,
of stable flow among hours (cf. above) can be identified in a and hence affected by hydropeaking. Our approach was based on
within-day hydrograph. Therefore, P denotes flow events lasting the comparison of the whole suite of STCM for pairs of similar river
between one and 24 h and which can be classified according to reaches, one with hydropeaking and the other free-flowing. We
the hourly pattern into periods of rise (RP), fall (FP), stability aimed at identifying the type of impact, as the attributes that were
(SP), minimum (MinP), and maximum (MaxP). STCM were devel- most impacted by the short-term flow regime, and the intensity
oped to quantify magnitudes, rates of change, frequencies, dura- and sign of the impact, as the degree of deviation from the refer-
tions and timing of P from each day of the year (i.e., ith day of ence condition and whether the impact increases or decreases
the year) (see Table 1 for detailed information on STCM). For whatever is being described by the metric. Type, intensity and sign
several-year long series (i.e., n years of the series), each metric of the short-term hydrological alteration are essential for under-
was computed as daily average for the whole dataset (Eq. (1)). standing its potential ecological consequences. We developed a
suite of Short-Term flow regime Impact Metrics (STIM) following
Xj¼n
STCMdayði;jÞ Eqs. (2), (3) and (4).
j¼1
STCMdayðiÞ ¼ ð1Þ Xi¼365
n STCMdayðiÞ
Consequently, there are 365 (or 366 for leap years) values for STCM ¼ i¼1
ð2Þ
365
each characterization metric. When flow records were available
for shorter time steps, such as 15-min intervals, they were trans- STCM dayðiÞðhpÞ
STCM dayðiÞðff Þ
If P0
formed into hourly flows by selecting only o’clock times. STCM STCMðff Þ
  ð3Þ
describing magnitude and rate-related features were previously STCM dayðiÞðhpÞ
STCM dayðiÞðff Þ
standardized by dividing between the mean hourly flow for the
STIM dayðiÞ ¼ Log 10 STCMðff Þ
þ1
dataset, which facilitates further comparisons with other river
reaches of different size. Similar to parameters defined by Richter STCM dayðiÞðhpÞ
STCM dayðiÞðff Þ
If <0
et al. (1996) for the characterization of long-term flow regime, STCMðff Þ
    ð4Þ
our STCM were assumed relevant for the biotic composition of STCM dayðiÞðhpÞ
STCM dayðiÞðff Þ
aquatic, wetland, and riparian ecosystems (Poff et al., 1997). Table 2 STIM dayðiÞ ¼ Log 10 STCM ðff Þ
þ1  ð
1Þ
briefly summarizes the main changes in the environment and
responses of organisms resulting from the alteration of the short- where hp and ff are the hydropeaking and free-flowing hourly flow
term flow regimes, which have been reported in previous studies. series, respectively. Each impact metric quantifies the deviation
Derived ecological consequences include reduced performance, from the reference condition of the corresponding characterization
slowed growth, and increased mortality of individuals, and ulti- metric (Eqs. (3) and (4)), and the intensity of the impact is relative
mately decreased species diversity, community composition shifts, to the yearly average of that metric in the reference conditions (Eq.
and eventually, exotic species invasion (see Table 2 for further (2)). Log10 was applied to the quotient to avoid excessively high val-
details and references). Such changes have been highlighted to ues when the yearly average of certain metrics in the reference con-
be very dependent on the species, age, life-stage and river mor- ditions are very low (e.g., metrics related to flow rates of change).
phology (Scruton et al., 2003; Roni et al., 2008; Tuhtan et al., Impact metrics can take any positive (Eq (3)) and negative value
2012; Hauer et al., 2014). (Eq (4)). Further details on STIM are shown in the Table 1.
Table 1 hourly flow records in the flow series from each site were interpo-
Description of the proposed Short-Term Characterization and Impact Metrics (STCM lated with the previous and following records.
and STIM).
We searched for significant differences among sites for the
Metric’s name, Metric’s definition characteristics and the alteration of their short-term flow regimes
abbreviation and units by running Kruskal-Wallis tests followed by the Games-Howell
Frequency Total periods of rise Daily mean total periods (GH) post hoc test on the sites’ STCM and STIM (P < 0.05 for signif-
(TRP; #periods/day) characterized by a sustained over icant results). Afterwards, regulated sites were classified based on
time hourly flow rise
Total periods of fall (TFP; Daily mean total periods
the impact of hydropeaking on their short-term flow regimes. On
#periods/day) characterized by a sustained over the one hand, we ranked the regulated sites based on the type
time hourly flow fall and intensity of the impact. For this, we averaged the absolute val-
Total periods of stability Daily mean total periods ues of the annually averaged impact metrics referring to all or each
(TSP; #periods/day) characterized by a sustained over
of the aspects of the flow regime and classified sites in four cate-
time hourly flow stability
Total periods of Daily mean total periods gories as slightly, moderately, highly, and strongly impacted. On
minimum (TMinP; characterized by a sustained over the other hand, we classified the sites per type, intensity, sign
#periods/day) time that day’s hourly minimum and timing of the impact of each site’s short-term flow regime.
flow For this, we carried out hierarchical cluster analysis using the Ward
Total periods of Daily mean total periods
method on the seasonally averaged impact metrics from each site
maximum (TMaxP; characterized by a sustained over
#periods/day) time that day’s hourly maximum for all or each of the flow regime aspects. Statistics were performed
flow in SPSS 23. We used Matlab 2015 for metric calculations.
Total hourly reversals Daily mean total rises and falls of
(TR; #reversals/day) hourly flows
Duration Duration periods of rise Daily mean duration of the periods 3. Results
(DurRP; h/day) characterized by a sustained over
time hourly flow rise
Duration periods of fall Daily mean duration of the periods 3.1. Characterization of short-term flow regimes
(DurFP; h/day) characterized by a sustained over
time hourly flow fall We defined 16 STCM (Table 1). Six metrics described frequency
Duration periods of Daily mean duration of the periods aspects of the short-term flow regime, and 5 described magnitude
stability (DurSP; h/day) characterized by a sustained over
and rate, and duration aspects, each. Information about timing was
time hourly flow stability
Duration periods of Daily mean duration of the periods also provided as each metric showed a value for each day of the
minimum (DurMinP; characterized by a sustained over year. Eleven out of the 16 metrics were uncorrelated (Supplemen-
h/day) time that day’s hourly minimum tary information S1). The number of periods of rise and fall (TRP
flow
and TFP), respectively, were positively correlated, so were the
Duration periods of Daily mean duration of the periods
maximum (DurMaxP; characterized by a sustained over
numbers of hourly reversals (TR) to both metrics, and the mean
h/day) time that day’s hourly maximum amplitude of hourly flows (MeanA) to the number of periods of
flow rise. Furthermore, the mean amplitude was negatively correlated
Magnitude Mean amplitude Standardized daily mean of the to the mean rise rate of hourly flows (MeanRR). Rise and fall
& rate (MeanA; unitless) difference between maximum and (MeanFR) rates of hourly flows were also negatively correlated.
minimum hourly flows Finally, the duration of periods of maximum and stability (Dur-
Mean minimum periods Standardized daily mean
MaxP and DurSP) were positively correlated. 1% of the flow data
of stability (MeanMinSP; minimum flow of periods
unitless) characterized by a sustained over were interpolated. Duration curves for the hourly flows and char-
time hourly flow stability acterization metrics highlighted the significant differences
Mean maximum periods Standardized daily mean between free-flowing and hydropeaking sites. The analyzed hydro-
of stability (MeanMaxSP; maximum flow of periods
power systems were designed to operate efficiently across the
unitless) characterized by a sustained over
time hourly flow stability
medium range of flows (i.e., around Q10%–Q80%), causing higher
Mean rise rate (MeanRR; Standardized daily mean hourly percentages of exceedance than in natural conditions (Fig. 3). They
unitless) flow rise rate are also able to cope and efficiently operate at higher flows (i.e.,
Mean fall rate (MeanFR; Standardized daily mean hourly Q1%–Q10%), up to the very rare extreme flood events (i.e., Q0%–
unitless) flow fall rate
Q1%) when water is spilled. Finally, they operate at a low power
output or begin to shut down at the lower flows (i.e., Q80%–
Q100%). Across these high and low flow ranges the operation
2.3. Data analysis results into lower percentages of exceedance than in natural con-
ditions (Fig. 3). During the optimum range of the plant operation,
STCM were computed for the 9-year hourly flow series from the most of the characterization metrics oscillated over a narrow range
11 selected sites to describe their short-term flow regimes. We con- of values which were higher than for natural flow regimes. This is
ducted an exploratory analysis of the hourly flow records and of the shown by the steeper slopes and upper position of the duration
STCM using duration curves and plotting the metrics against flows. curves of the metrics from the hydropeaking sites between
Additionally, we looked for the correlated metrics as those with STCM10%–STCM80% (Fig. 4), as well as by the unscattered cloud
Spearman’s Correlation Index higher than ±0.7. Based on the STCM, of hydropeaking metrics values above those from the free-
the series of STIM was computed for the eight sites at the Ume River flowing sites (Supplementary information S2). Some duration-
to evaluate the type, intensity and sign of short-term alteration related metrics, however, were lower in hydropeaking than in
resulting from hydropeaking. According to their hydrological and free-flowing regimes from a certain percentage of exceedance of
physical characteristics, Sorsele and Granåker sites in the Vindel the metric (e.g., DurRP or DurFP; Fig. 4). Additionally, unlike in
River were selected as references for the upper site (i.e., Grundfors) the free-flowing sites, the metrics’ lowest values (i.e., above
and for the remaining sites (i.e., from upstream to downstream: STCM80%) in the hydropeaking sites did not equal zero for most
Rusfors, Bålforsen, Betsele, Tuggen, Bjurfors övre, Bjurfors nedre, cases and the majority of them were exceeded almost 100% of
and Harrsele) in the Ume River, respectively (Fig. 1). Non-existent the time (Fig. 4).
Table 2
Summary of the main changes in the environment and responses of organisms resulting from the alteration of the short-term flow regimes, which have been reported in
representative studies.

Flow feature and alteration Changes on environment and responses of organisms References
Frequency Increased within-day number of Fish behavioral changes including very active fish which show many Scruton et al. (2003), Taylor et al. (2014),
reversals non-migratory short movements (e.g., constant change from resting on Capra et al. (2016), Boavida et al. (2017)
substrate to swimming in the water column, and search for suitable
habitats). This implies higher fish energy consumption.
Fish habitat persistence undermined, involving transient shelters and Freeman et al. (2001), Flodmark et al.
food, which increases competence with neighboring fish and reduces (2004)
food uptake.
Increased scouring capabilities of moving water and ice, which damages Jensen and Johnsen (1999), Friedman and
or removes sessile organisms or life stages, such as fish eggs, aquatic Auble (1999), Madsen et al. (2001), Lind
macroinvertebrates and aquatic and riparian plants. et al. (2014)
Increased turbidity, which affects adversely fish movement and health Bruton (1985), Kirk (1994)
by clogging gill rakers and gill filaments, and hampers aquatic and
riparian plants photosynthesis.
Increased within-day number of High prey (e.g., invertebrates) availability, which results in more food Rocaspana et al. (2016), Kelly et al. (2017)
periods of flow rise, and of for fish.
maximum flows Frequent inundation of riparian areas causing soil waterlogging and Ernst (1990), Armstrong et al. (1994)
frequent partial or total submergence. This affects aquatic and riparian
plants by attenuating light, hampering gas exchange and causing
anoxia.
Fish migratory changes such as increased upstream migration. Taylor and Cooke (2012)
Magnitude Increased within-day Fish habitat deterioration or improvement. Deterioration through Jensen and Johnsen (1999)
magnitude of periods of substrate mobilization; improvement through maximizing oxygen
maximum flows supply when promoting flushing of fine sediment from gravel.
Fish egg damage and removal. Jensen and Johnsen (1999), Gostner et al.
(2011)
Macroinvertebrate removal. Gostner et al. (2011), Miller and Judson
(2014)
Increased scouring capabilities of high flows, which damages or Friedman and Auble (1999), Madsen et al.
removes sessile organisms or life stages, such as fish eggs, (2001)
macroinvertebrates and plants directly or indirectly through soil erosion
and changes in channel morphology.
Decreased within-day Reduction of suitable habitats available for fish due to lower volume of Vollset et al. (2016)
magnitude of periods of water in the river and reduction in average depth and width of the river
minimum flows channel, which result in oxygen stress and cause problems with fish
refuge and feeding.
Habitat deterioration following siltation of gravel and reduced oxygen, Cowx et al. (1998), Levasseur et al. (2006)
which affects fish spawning.
Fish migratory changes both upstream and downstream due to too little Armstrong et al. (2003)
water that impedes fish movement
Elevated fish density due to inability to redistribute themselves. Fish Smith and Reay (1991)
restricted to wetted ponds, which may lead to cannibalism, limited
feeding, and higher transmission of pathogens
Fish egg desiccation or freezing (in winter). Milner et al. (2003), Casas-Mulet et al.
(2015)
Indirect effects on fish of increased water temperature Elliott et al. (1997), Wootton (1998)
Desiccation of macroinvertebrates. Holzapfel et al. (2016)
Soil moisture deficit and water stress for aquatic and riparian plants. Garcia de Jalón et al. (1998), Porporato et al.
(2001), Stella et al. (2010), Mjelde et al.
(2013)
Rates Rapid within-day flow decrease Fish stranding. Saltveit et al. (2001), Irvine et al. (2009),
Tuhtan et al. (2012), Hauer et al. (2014),
Boavida et al. (2015)
Deposition of plant propagules and soil. Seeds and vegetative parts on Goodson et al. (2003), Pettit and Naiman
newly deposited surfaces favor plant establishment, but soil deposition (2006)
also results into coating mud, burial, and soil surface clogging.
Hampered germination. Baldwin et al. (2001)
Rapid within-day flow increase Fish drift. Wolter and Sukhodolov (2008)
Macroinvertebrate drift (catastrophic drift). Timusk et al. 2016; Bruno et al. 2016;
Leitner et al. (2017), Gostner et al. (2011)
Increased aquatic and riparian plant propagules dispersal. Boedeltje et al. (2004)
Hampered germination. Baldwin et al. (2001)
Duration Shorter within-day periods of That such periods are shorter implies less time for key activities of fish Capra et al. (2016), Boavida et al. (2017)
flow rise and fall such as feeding in the water column or resting on substrate, which
ultimately results in less food uptake and higher fish energy
consumption.
Longer within-day periods of That such periods are longer exacerbates their above-described effects. Warren et al. (2015), Casanova and Brock
minimum and maximum flows For example, likelihood of fish egg desiccation and death following (2000)
stranding increases with duration of low flow periods. Similarly,
likelihood of death of riparian or aquatic plants increases when they
remain under water or emerged for a long period, respectively.

(continued on next page)

Table 2 (continued)

Flow feature and alteration Changes on environment and responses of organisms References
Timing Seasonality of short-term flow Increased fish drift and stranding risk during winter. Saltveit et al. (2001), Halleraker et al. 2003;
changes Irvine et al. (2015)
Increased fish migratory and non-migratory behavioral changes during Puffer et al. (2015)
summer.
Freezing of eggs during winter.
Low flows naturally taking place in winter reduce drift and favor benthic Heggenes et al. (2016)
feeding, which is more profitable than drift feeding due to the lower cost
of not having to maintain a position in the current. Getting the same
amount of food under higher winter drift implies more energy
expenditure.
Timing of life stages of aquatic insects and aquatic and riparian plants Castro et al. (2013), Kennedy et al. (2016)
determines how they are affected by short-term flow changes. For
example, short-term flow alterations coinciding with adult egg-laying
behavior such that open-water layers or with plants dispersal will not
have major effects, whereas alterations coinciding with river-edge
layers, such as mayflies, or with seed germination will cause severe
impact.

Granåker day than the free-flowing ones (Fig. 5). The frequency of hourly
100 reversals per day (TR) was remarkably higher in the hydropeaking
Sorsele
90 Gautsträsk sites than in the free-flowing sites (Fig. 5). Despite being more fre-
Percentage exceedance (%)

80 Harrsele quent, periods of rise and fall per day were considerably shorter in
B.övre the hydropeaking sites compared to the free-flowing sites (Fig. 5).
70
B.nedre Periods of stability were of similar duration for the downstream
60 hydropeaking and free-flowing sites and slightly longer for the
Tuggen
50 Betsele most upstream hydropeaking sites (Fig. 5). There were no differ-
40 Bålforsen ences in the daily frequencies of the periods of minimum and max-
Rusfors imum (TMinP and TMaxP) between free-flowing and hydropeaking
30
Grundfors sites, and their durations were slightly different. Most magnitude
20 metrics showed higher dispersion in the free-flowing sites
10 (Fig. 5). Whereas the median for the mean daily hourly amplitude
0 (MeanA), and for the daily periods of flow stability’s maximum
0 1 2 3 4 5 value (MeanMaxSP) was significantly higher along hydropeaking
Standardized hourly flow sites (Fig. 5), the median for the minimum daily value of flow sta-
bility (MeanMinSP) was significantly lower in the hydropeaking
Fig. 3. Flow duration curve for hourly flows recorded at each studied site. Grey lines sites compared to the free-flowing ones. Hourly flows rose and fell
represent the free-flowing sites, whereas the black lines represent the hydropeak-
(MeanRR and MeanFR) significantly faster under hydropeaking
ing sites. Note that the maximum standardized flows from the free-flowing sites are
20.1 (Gautsträsk), 10.0 (Sorsele) and 7.9 (Granåker), but for representation conditions (Fig. 5).
purposes, the x-axis only contains standardized flows up to 5.

3.2. Assessment of short-term flow regimes alteration

All characterization metrics were significantly different
between sites (p < 0.05; Fig. 5; Supplementary information S3). There were 16 STIM, each corresponding to a STCM (Table 1).
Post-hoc tests revealed that all characterization metrics differed Impact metrics ranged between
0.5 and 2 (Fig 6). Between sites
significantly between the free-flowing and hydropeaking sites (dif- comparisons of the impact metrics showed the following results.
ferences of the metric values are seen rightward and leftward of Rates were similarly impacted in all sites (i.e., P > 0.05 for
the vertical dashed line in Fig. 5). However, some metrics were MeanRR-I and MeanFR-I; Fig. 6; Supplementary information S5).
similar for free-flowing and hydropeaking sites, such as TSP, Among significantly different impact metrics (P < 0.05), some dura-
TMinP, DurMinP, DurMaxP and MeanMaxSP (Fig. 5; Supplemen- tion related metrics only differed in one or two sites such as DurFP-
tary information S3). In addition, some metrics differed signifi- I and DurMaxP-I, whereas others differed in many of the sites (TRP-
cantly within the pool of hydropeaking sites (leftward of the I, TFP-I, TR-I, DurMinP-I, MeanA-I, and MeanMinSP-I; Fig. 6 and
dashed line) or within the pool of free-flowing sites (rightward of Supplementary information S5). Seasonally averaged impact met-
the dashed line; Fig. 5). For example, most of the sites showed sig- ric values are detailed in Supplementary information S6. Below fol-
nificantly different values for TFP, DurFP, MeanA and MeanRR, lows a detailed evaluation of the impacts on the different metrics.
whereas MeanMinSP was significantly higher in the free-flowing Frequency related features were in general moderately
sites compared to the hydropeaking sites but barely differed impacted (Fig 6). As shown by the relatively high positive value
among the hydropeaking ones (Fig. 5). As shown by the width of of the impact on the number of periods of rise and fall and hourly
the boxes in Fig. 5, in general, metrics characterizing the free- reversals per day (TRP-I, TFP-I and TR-I), these events tended to be
flowing sites varied along the year more than in the hydropeaking much more frequent in the sites subjected to hydropower produc-
sites (Fig. 5). Seasonally averaged characterization metrics values tion than in the free-flowing ones, being the impact higher as we
are detailed in Supplementary information S4. move downstream (Fig. 6). The seasonal analysis of these metrics
As general patterns, most hydropeaking sites showed more also highlights that the impact was most evident during autumn
periods of rise and fall (TRP and TFP) and less periods of stability and winter (Fig. 6). The frequency of periods of stability per day
(TSP) per day than the free-flowing sites. Some upstream (TSP-I) was impacted oppositely in the downstream and upstream
hydropeaking sites, however, were similarly or more stable per sites. In general, whereas it slightly decreased downstream and
 TRP Hydropeaking TR

Percentage exceedance (%)

Percentage exceedance (%)
100 (e.g., Harrsele) 100
a) Free-flowing b)
80 (e.g.,Granåker) 80

60 60

40 40

20 20

0 0
0 1 2 3 4 0 2 4 6 8 10 12

TRP (#periods/day) TR (#reversals/day)

DurRP DurFP

Percentage exceedance (%)

Percentage exceedance (%)

100 100
c) d)
80 80

60 60

40 40

20 20

0 0
0 5 10 15 20 25 0 5 10 15 20 25

DurRP (h/day) DurFP (h/day)

MeanA MeanRR
Percentage exceedance (%)
Percentage exceedance (%)

100 100
e) f)
80 80

60 60

40 40

20 20

0 0
0 0.5 1 1.5 2 0 0.1 0.2 0.3
Stand. MeanA (unitless) Stand. MeanRR (unitless)

Fig. 4. Duration curve for a selection of Short-Term Characterization Metrics in a selection of two sites, one free-flowing (Granåker; dashed grey line) and the other subjected
to hydropeaking (Harrsele; solid black line). Abbreviations, units and other information on the STCM are detailed in Table 1.

mainly during summer and autumn, it moderately increased For example, hydropeaking increased moderately the maximum
upstream under hydropeaking conditions. No particular seasonal hourly flows during autumn and winter, whereas it hardly
pattern was found for the TSP-I along the upstream sites (Fig. 6). decreased them during spring and summer. In contrast,
Finally, the impact on the frequencies of the periods of minimum hydropeaking moderately decreased the minimum hourly flows
and maximum per day was very low (TMinP-I and TMaxP-I mean and strongly increased the hourly rise rates in all seasons but win-
close to zero) for all sites (Fig. 6). Impact on the duration related ter (Fig. 6). The impact on the mean amplitude and the mean
features was the lowest compared to other flow features (Fig. 6). hourly fall rates (MeanA-I and MeanFR-I) was strong regardless
Depending on the metric, the site and the season, hydropeaking of the season, and except for the MeanA-I, which was slightly more
resulted in shorter or longer hourly events than in natural condi- affected downstream, all magnitude and rate related metrics were
tions (Fig. 6). Events of rise and fall (DurRP-I and DurFP-I) were similarly impacted in all sites (Fig. 6).
moderately shorter in all sites affected by hydropeaking than in
the free-flowing ones, but the highest impacts occurred during 3.3. Classification of sites based on short-term alteration of flow
spring for the rise events, and during autumn and winter for the regimes
fall events (Fig. 6). The daily duration of the periods of stability,
minimum and maximum (DurSP-I, DurMinP-I and DurMaxP-I) When averaging the absolute values for the annually averaged
hardly differed between hydropeaking and free-flowing conditions, STIM (for all metrics or for each subset referred to each flow fea-
and in general, minor changes on these features occurred predom- ture), the intensity of the impact ranged between 0 and 1.05. Hence,
inantly during winter and spring (Fig. 6). Finally, impact on the the simplest classification of the sites was based on the impact type
magnitude and rate related features was the highest (Fig. 6). The and intensity, and regardless of the sign and the timing, and con-
impact on the magnitude of the flows characterizing periods of sisted of four categories: free-flowing (0), slightly impacted
daily maximum and minimum and of the hourly rise rates (<0.25), moderately impacted (0.26–0.51), highly impacted (0.52–
(MeanMinSP-I, MeanMaxSP-I and MeanRR-I) varied with seasons. 0.77), and strongly impacted (0.78–1.05; Fig. 7). The flow regime
 6 a) TRP b) TFP c) TSP 12 d) TR
10
4 8
6
2 4
2
0 0

3 e) TMaxP f) TMinP 24 g) DurRP h) DurFP

18
2
12
1
6
0 0

24 i) DurSP j) DurMaxP k) DurMinP 3 l) MeanA

18
2
12
1
6
0 0

3 m) MeanMaxSP n) MeanMinSP .4 o) MeanRR 0 p) MeanFR

.3 -.1
2
.2 -.2
1
.1 -.3
0 0 -.4
H N O T B L R G U S K H N O T B L R G U S K H N O T B L R G U S K H N O T B L R G U S K

Fig. 5. Box-and-whisker plots for the daily averaged values of the Short-Term Characterization Metrics (STCM) for studied sites. Graphs show the median (dark horizontal
line), percentiles 75 and 25 (edges of boxes), percentiles 90 and 10 (end of vertical lines), and outliers (asterisks). Dashed vertical lines on the top of graphs divide the free-
flowing sites towards the right and the perturbed sites towards the left. Abbreviations, units and other information on the STCM are detailed in Table 1. Abbreviations of sites
are detailed in Table 3. Graphs a– f show the frequency related metrics, g– k show the duration related metrics, and l– p show the magnitude and rate related metrics.

in Grundfors (most upstream site) was least impacted for any of the Black et al. 2005; Moliere et al. 2009; Bejarano et al. 2010;
features analyzed. It was followed by Rusfors, whose short-term Kennard et al. 2010; McManamay et al. 2012), few methods adopt
flow regime was moderately impacted in terms of frequencies data at such a high resolution that is necessary for understanding
and durations, but strongly impacted in terms of magnitudes and the within-day hydrograph. Despite short-term flow analysis being
rates. The short-term flow regimes of remaining sites were strongly viable with shorter hydrological series than those typically needed
impacted for magnitudes and rates, relatively highly impacted for to assess seasonal or annual flow patterns (Bevelhimer et al. 2015),
frequencies, and moderately for durations (Fig. 7). Finally, the sea- the limited availability of hourly and sub-hourly stream flow
sonality of the hydropeaking impact was corroborated by the most records and the laborious processing required by such large volume
comprehensive classification of the sites, i.e., the hierarchical clas- of data have been a handicap for studies at finer resolutions. More-
sification. Between three and five groups resulted from the dendro- over, although government agencies from several countries and
grams (distance 4) according to the type, intensity and sign of the hydropower companies are recently making their instantaneous
impact on each season’s short-term flow regime (Table 3; Supple- flow records available [e.g., the USGS Current Water Data for the
mentary information S7). This classification showed that, except Nation (EEUU); the Swedish Meteorological and Hydrological Insti-
for the frequency related metrics, the impact strongly varied with tute and some Swedish hydropower companies; the Norwegian
season (Table 3). In general, from the lowest to the strongest Water Resources and Energy Directorate; the Rete di Monitoraggio
impact, groups based on the impact on frequencies distinguished in Tempo Reale dell’Ufficio Dighe (Italy); the Federal Office for the
between sites regardless of the season, i.e., Grundfors, Rusfors, Environment (BAFU; Switzerland), among others], there is still a
and the remaining downstream sites. Groups based on the impact lack of tools supporting their treatment (but see, for example,
on magnitudes and on all features together distinguished between Sauterleute and Charmasson 2014; Haas et al. 2014). For our pur-
seasons regardless of the sites, i.e., in general, the short-term flows pose, the Vindel and Ume rivers in northern Sweden were excellent
from any site in winter, autumn, spring and summer. Finally, groups study areas. The long series of sub-daily flow records available pro-
based on the impact on durations combined sites and seasons vided a wide range of scenarios of short-term flow patterns, as they
(Table 3; Supplementary information S7). came from free-flowing sites (the Vindel River) and from sites used
for hydropower production (the Ume River). Furthermore, as both
4. Discussion rivers originally had similar hydrological and physiographical char-
acteristics, the space-by-time method is viable for alteration assess-
Whereas many studies up to date have dealt with seasonal and ment of the Ume River sites, using the Vindel River as reference. In
annual flow patterns (Clausen and Biggs, 2000; Harris et al. 2000; fact, this pair of rivers has been successfully used in several
 2.0
a) TRP-I b) TFP-I c) TSP-I d) TR-I
1.5

1.0

0.5

0.0

-0.5
2.0
e) TMaxP-I f) TMin-I g) DurRP-I h) DurFP-I
1.5

1.0

0.5

0.0

-0.5
2.0
i) DurSP-I j) DurMaxP-I k) DurMinP-I l)
1.5

1.0

0.5

0.0
MeanA-I
-0.5
2.0
m) MeanMaxSP-I n) MeanMinSP-I o)
1.5
(p)
1.0

0.5

0.0
MeanRR-I MeanFR-I
-0.5
H N O T B L R G H N O T B L R G H N O T B L R G H N O T B L R G

Fig. 6. Bar plots showing the seasonally averaged Short-Term Impact Metrics (STIM) for each study site. Black bars represent the autumn, grey bars represent the winter, light
grey bars represent the spring, and white bars represent the summer. The x-axis contains the perturbed sites whose abbreviations are detailed in the legend of Table 3. The y-
axis contains the metrics which are unitless and whose abbreviations are detailed in Table 1. Graphs a–f show the frequency related metrics, g– k show the duration related
metrics, and l–p show the magnitude and rate related metrics.

previous comparative studies (Nilsson et al., 1991; Merritt et al., stability trends of flow, or by extreme (maximum or minimum)
2010). The data treatment tool we developed also speeded up met- flows. All of them have been proved to be key for functions of
ric calculations. Nevertheless, while short-term flows are usually the ecosystem. For example, rapid dewatering was shown to
recorded at hydropower plants, we are aware of that recorded data increase the stranding risk of adult fish resulting in higher mortal-
from comparable pristine conditions are not commonly found else- ities (Saltveit et al., 2001; Scruton et al., 2005 and Scruton et al.,
where. In these cases, applying the approach would involve the 2008), increase riparian plant seedling mortality (Stella et al.,
restitution of the natural short-term flow regime at the plant loca- 2010) and slow down their growth (Amlin and Rood, 2002), and
tion, or the generation of flows at the desired short-term resolution favor sandbar erosion ultimately impacting biological communi-
for a long period based on commonly available daily flows ties because of habitat loss (Álvarez and Schmeeckle, 2013). In
(recorded or modeled) and assuming similar within-day flow vari- addition, high daily flow fluctuations were shown to decrease adult
ability than the one from, at least, a representative year of recorded fish size (Dibble et al., 2015), whereas increases in frequency, dura-
data. Minute- or hourly-flows and levels may be measured rela- tion and depth of daily inundations resulted in a reduction of
tively cheap and easy using pressure-transducer loggers. diversity of plant communities by hindering seed germination
One of the common examples of short-term flow regime alter- (Sarneel et al., 2014) and seedling growth (Baldwin et al., 2001).
ations is hydropower production through peaking plants. The The magnitude and duration of water releases from the power
recent increase in hydropower production worldwide has triggered plant also caused considerable losses from benthic populations to
an interest in short-term flow regimes, as these are highly affected drift (Bruno et al., 2010) (See literature reviewed in Table 2). Con-
by hydropeaking and in recognition of their influence on fluvial sequently, all devised metrics are worth keeping as long as they all
ecosystems. However, many studies still have a narrow scope deal- have ecological importance. Given that our proposed methodology
ing with a small and biased group of metrics (Meile et al., 2011; may be potentially used in any context of short-term flow regime
Zimmerman et al., 2010; Carolli et al., 2015), as long as they are characterization and impact assessment, even metrics that hardly
essential for a target species, usually fish (Halleraker et al., 2003), differed between our testing sites might be significantly altered
or include larger but still insufficient numbers to account for every in other cases.
ecologically relevant aspect of the sub-daily flow regime (Haas The productivity limitations of hydropower plants (in accor-
et al., 2014; Sauterleute and Charmasson, 2014; Bevelhimer dance with, e.g., the characteristics of the turbines, dam height or
et al., 2015; Chen et al., 2015). Our proposed metrics have been reservoir storage capacity), together with the variable demand
carefully thought out to enable capturing the full range of for hydroelectricity (in accordance with, e.g., the availability of
within-day hydrological facets which may potentially influence other energy sources and with daytime and season) (source: U.S.
any of the elements (biotic and abiotic) of the fluvial ecosystem, Energy Information Administration; http://www.eia.gov/) result
at any time of the life-cycle of different individuals. These metrics into highly variable operation schemes (i.e., forms of hydropeak-
provide information on frequency, duration, magnitude, rate of ing), involving different intensities of production and frequencies
change and timing of daily events characterized by rise, fall or and durations of periods of production over time. Furthermore,
 (a)

¯
(b)

(c) (d)

Free-flowing
Slightly impacted
Moderately impacted
Highly impacted
Strongly impacted

60 km

Fig. 7. Ranking of the hydropeaking sites based on (a) the intensity of the impact on all, or (b) on each of the aspects of the flow regime (b: frequency; c: duration; d:
magnitude and rate related metrics). The data shown on the maps are averages of the absolute values for the annually averaged STIM. Colors represent different impact
intensities: grey (=0), blue (<0.25), yellow (0.26–0.51), orange (0.52–0.77), and red (0.78–1.05). Sites can be consulted in Fig. 1.

temporal patterns in power plant operation are superimposed on in the growing season where this attribute is particularly altered
the natural temporal variability exhibited by the flow regimes of (Gorla et al., 2015). This would be the case of our studied upstream
some river types. Consequently, similar operating rules may result regulated sites, where hydropeaking leads to slightly longer
in different impacts on flows depending on, e.g., daytime, weekday within-day periods of high flows. Other examples involve defining
or season. Therefore, an added value of our devised metrics is that mitigation measures to impede fish stranding, which has been
they are able to discern not only which flow features, but also to widely documented as the main reason for fish mortality along
what degree, how and when they are altered, being able to accu- reaches affected by hydropeaking (see reviews by Nagrodski
rately characterize and evaluate any form of hydropeaking. Budget et al., 2012 and Irvine et al., 2015). As observed in many studies
constraints may lead us to focus mitigation measures on a single, (Saltveit et al., 2001; Scruton et al., 2005, 2008), the response of
the most affected flow attribute. Alternatively, one may be only fish to remain relatively sedentary in winter under varying flows,
interested in restoring a flow attribute because it is key for a target may increase the likelihood for dewatering, stranding and freezing
species which is particularly sensitive at a specific moment of its leading to higher mortality. In winter, energy reserves are depleted
life cycle. Therefore, consideration of impact type, intensity, sign and any activity and/or stress related to hydropeaking may poten-
and timing helps targeting of management, making it more feasible tially affect production and survival. River morphology also plays a
and cost-effective. For example, operational mitigation measures relevant role as stranding risk has been recently reported higher
involving flow management to ensure proper growth of riparian along braided compared to single-thread rivers (Vanzo et al.,
seedlings should be focused on decreasing flooding duration early 2015). Mitigation proposals to favor fish populations in the studied
 Table 3
Summary of groups resulting from the cluster classification of the river sites subjected to hydropower production (H: Harrsele; N: Bjurfors nedre; O: Bjurfors övre;
T: Tuggen; B: Betsele; L: Bålforsen; R: Rusfors; G: Grundfors) according to their seasonally averaged Short-Term Impact Metrics. Table shows groups from
dendrograms at a selected distance 4. Classifications were based on the pool of all metrics regardless the hydrological feature they refer to (first column), and on
the sub-set of metrics referring exclusively to frequency (second row), duration (third row), and magnitude and rate (fourth row) related features. Dendrograms
may be checked in Supplementary information S7.

Classification according to impact Group Sites Seasons

on metrics related to:
Frequency, duration, magnitude and rates (i) H, N, O, T, B, L Spring and summer
(ii) H, N, O, T, B, L Autumn
(iii) H, N, O, T, B, L, R Winter
(vi) R Autumn, spring and summer
G Spring and summer
(v) G Autumn and winter
Frequency (i) H, N, O, T, B, L, R, G Autumn, winter, spring and summer
(ii) R Autumn, winter, spring and summer
(iii) G Autumn, winter, spring and summer
Duration (i) B, R Autumn, winter and summer
G Summer
(ii) H, N, O, T, B, L, R, G Spring
(iii) H, N, O, T, L Winter
(iv) H, N, O, T, L Autumn and summer
G Autumn and winter
Magnitude and rates (i) H, N, O, T, B, L, R, G Spring and summer
(ii) H, N, O, T, B, L, R, G Autumn
(iii) G Winter
H, N, O, T, B, L, R Winter

reach from the Ume River should involve decreasing numbers of in general, operational mitigation measures could be similar for all
fall events and flow fall rates during winter mainly along the more power plants downstream of Bålforsen only differing depending on
morphologically complex downstream sites. Magnitude alterations the season, but should be specifically adapted to each of the two
of sub-daily flow events may be important for sessile species, such remaining upstream plants (i.e., Grundfors and Rusfors) being the
as plants or macroinvertebrates. Riparian species naturally grow- same during the whole year. Management groups of sites, how-
ing along the upper riparian zone experience sporadic flooding, ever, would be others if certain flow features were the objective
whereas aquatic and amphibious plants growing further down in of the management action. The goal of the approach is to be useful
the riparian zone face almost constant inundation (Colmer and for the sustainable water management planning, which must con-
Voesenek, 2009). Wide ranges of within-day flows, as in rivers sub- sider flow variability at time scales from minutes to decades, as
jected to hydropeaking, expose aquatic and riparian-terrestrial time is widely recognized as the fourth dimension influencing flu-
plants to frequent variation between drainage and flooding, which vial ecosystems (Ward, 1989; Poff et al., 1997; Biggs et al., 2005). It
impacts riparian areas (Havens et al,. 2004; Bailey-Serres and may be particularly helpful in river basins used for hydropower
Voesenek, 2008). This effect is particularly strong in the down- production. Studies like this can inform decision making on restor-
stream sites of our study, where measures aimed at narrowing ing river reaches affected by hydropeaking, relicensing or increas-
within-day flow ranges are expected to benefit riparian ing the capacity of already existing hydropower plants such as in
ecosystems. Europe or North America, or evaluating the convenience of new
Existing approaches aimed at evaluating hydropeaking impacts developments as it is the case of Asia, Africa and South America.
on sub-daily flow regimes only consider a few metrics (Carolli For example, the economic benefits from hydropower generation
et al., 2015). Others were primarily developed for characterization along the Mekong or Amazon rivers should be carefully evaluated
of short-term flows, and although they might be suitable also for against the potentially irreversible loss of ecosystem services such
impact assessment, this is not straightforward (Zimmerman as loss of fisheries, removal of wetlands and loss of inundated for-
et al., 2010; Meile et al., 2011). More comprehensive approaches ests, or effects on food security (Grumbine and Xu, 2011; Lees
such as those of Sauterleute and Charmasson (2014) and Chen et al., 2016). This begins by analyzing the impacts of different pro-
et al. (2015) are well suited to describe short-term flow regimes duction (i.e., economic) scenarios on flows at all time scales, and
affected by hydropeaking, but they are inadequate for detecting continues with the evaluation of their ecological and societal con-
alterations. Furthermore, neither of these methods can quantify sequences. For the former, our approach would complement stud-
the various temporal aspects of variability. Only the Wavelet ies on daily, monthly or seasonal basis (Räsänen et al., 2012;
Transform approach is applicable to assess alterations with time, Kuenzer et al., 2013). For the latter, further research should provide
but emphasizes the identification of impacted time scales (White knowledge on relationships between sub-daily flows and ecology
et al., 2005). Our proposed approach overcomes these drawbacks (and society). Overall, harmonizing hydroelectricity production
by characterizing both free-flowing and perturbed sub-daily flows and fluvial ecosystem sustainability is a challenging demand for
and assessing their impact with time, through several ecologically industry and society, and studies like this are absolutely needed.
meaningful metrics. In addition, metrics may be easily transformed
to stage when the input is a stage series recorded every hour or
less. This is interesting when investigating the effects of 5. Conclusions
hydropeaking on riparian ecosystems, given that plants are sessile
organisms fully dependent on water-level fluctuations. Finally, The approach presented in this article, which encompasses
proposed classifications help identify sites which have the same devising new characterization metrics and the adaptation of eval-
short-term flow regime alterations and, consequently, similar uation and classification methods standardly used in ecology, rep-
management might be recommended for the entire group. Accord- resents a breakthrough in the still poorly studied field of short-
ing to our results from the most comprehensive site classification, term flows. It has been conceived to support the sustainable water
management in rivers subjected to hydropower production, and in Bevelhimer, M.S., McManamay, R.A., O’Connor, B., 2015. Characterizing sub-daily
flow regimes: implications of hydrologic resolution on ecohydrology studies.
general in rivers whose sub-daily flows are altered by any other
River Res. Appl. 31, 867–879.
cause. It may also help directing measures of fluvial restoration Biggs, B.J., Nikora, V.I., Snelder, T.H., 2005. Linking scales of flow variability to lotic
to particularly impacted river reaches, flow features and seasons. ecosystem structure and function. River Res. Appl. 21, 283–298.
It could also serve public authorities during flow regulation re- Black, A.R., Rowan, J.S., Duck, R.W., Bragg, O.M., Clelland, B.E., 2005. DHRAM: a
method for classifying river flow regime alterations for the EC Water
licensing or evaluating processes of new projects. Site-specificity Framework Directive. Aquat. Conserv.-Mar. Freshwater Ecosyst. 15, 427–446.
of water-stage results in complex relationships between this vari- Boavida, I., Santos, J.M., Ferreira, T., Pinheiro, A., 2015. Barbel habitat alterations due
able and hydropeaking. Therefore, future research should address to hydropeaking. J. Hydro-environ. Res. 9, 237–247.
Boavida, I., Harby, A., Clarke, K.D., Heggenes, J., 2017. Move or stay: habitat use and
characterization, impact assessment and classification of short- movements by Atlantic salmon parr (Salmo salar) during induced rapid flow
term water-level regimes, which are key when analyzing potential variations. Hydrobiologia 785, 261–275.
impacts of hydropower production on riparian zones. In addition, Boedeltje, G.E.R., Bakker, J.P., Ten Brinke, A., Van Groenendael, J.M., Soesbergen, M.,
2004. Dispersal phenology of hydrochorous plants in relation to discharge, seed
our approach should be extended to other river reaches worldwide, release time and buoyancy of seeds: the flood pulse concept supported. J. Ecol.
as different operating schemes on different river types may lead to 92, 786–796.
different impact ranges from those obtained here. High impacts in Bruno, M.C., Maiolini, B., Carolli, M., Silveri, L., 2010. Short time-scale impacts of
hydropeaking on benthic invertebrates in an alpine stream (Trentino, Italy).
snowmelt-fed rivers are due to the natural stability of stream Limnologica 40, 281–290.
flows, but we might expect much lower impacts in, e.g., Mediter- Bruno, M.C., Cashman, M.J., Maiolini, B., Biffi, S., Zolezzi, G., 2016. Responses of
ranean rivers given their great dynamics even under natural condi- benthic invertebrates to repeated hydropeaking in semi-natural flume
simulations. Ecohydrology 9, 68–82.
tions (Gasith and Resh, 1999). A global application would also
Bruton, M.N., 1985. The effects of suspended solids on fish. Hydrobiologia 125, 221–241.
generate a global database of sub-daily flow-regime types and Capra, H., Plichard, L., Bergé, J., Pella, H., Ovidio, M., McNeil, E., Lamouroux, N., 2016.
impacts providing objective information necessary for categoriza- Fish habitat selection in a large hydropeaking river: strong individual and
temporal variations revealed by telemetry. Sci. Total Environ. 578, 109–120.
tion of river reaches worldwide. Finally, ecological implications
Carlisle, D.M., Wolock, D.M., Meador, M.R., 2011. Alteration of streamflow
of described sub-daily flow changes still need to be sorted out magnitudes and potential ecological consequences: a multiregional
(Poff et al., 2010). assessment. Front. Ecol. Environ. 9, 264–270.
Carolli, M., Bruno, M.C., Siviglia, A., Maiolini, B., 2012. Responses of benthic
invertebrates to abrupt changes of temperature in flume simulations. River Res.
Acknowledgments Appl. 28, 678–691.
Carolli, M., Vanzo, D., Siviglia, A., Zolezzi, G., Bruno, M.C., Alfredsen, K., 2015. A
simple procedure for the assessment of hydropeaking flow alterations applied
The authors would like to thank the Swedish Meteorological to several European streams. Aquat. Sci. 77, 639–653.
Agency -SMHI- and Statkraft Sverige AB, Vattenfall AB, E.ON Vat- Casanova, M.T., Brock, M.A., 2000. How do depth, duration and frequency of
tenkraft Sverige AB, Vattenfall Vattenkraft AB and Vattenfall Umeäl- flooding influence the establishment of wetland plant communities? Plant Ecol.
147, 237–250.
ven AB power companies, for providing the series of instantaneous
Casas-Mulet, R., Saltveit, S.J., Alfredsen, K., 2015. The survival of Atlantic salmon
flows. Roland Jansson is gratefully acknowledged for his valuable (Salmo salar) eggs during dewatering in a river subjected to hydropeaking. River
comments during the study process. The research leading to these Res. Appl. 31, 433–446.
Castro, D.M., Hughes, R.M., Callisto, M., 2013. Effects of flow fluctuations on the
results was funded by the EU-FP7-PEOPLE Marie Curie IEF actions
daily and seasonal drift of invertebrates in a tropical river. Ann. Limnol. Int. J.
under the grant agreement no 623691 (RiPeak: ‘‘Responses of Limnol. 49, 169–177.
RiParian forests to hydroPeaking: towards a sustainable hydro- Chen, Q., Zhang, X., Chen, Y., Li, Q., Qiu, L., Liu, M., 2015. Downstream effects of a
power management”; http://cordis.europa.eu/project/rcn/187789_es. hydropeaking dam on ecohydrological conditions at sub-daily to monthly time
scales. Ecol. Eng. 77, 40–50.
html). Funds from Fundación José Entrecanales Ibarra also Clausen, B., Biggs, B.J.F., 2000. Flow variables for ecological studies in temperate
supported part of the research. streams: groupings based on covariance. J. Hydrol. 237, 184–197.
Colmer, T.D., Voesenek, L.A.C.J., 2009. Flooding tolerance: suites of plant traits in
variable environments. Funct. Plant Biol. 36, 665–681.
Appendix A. Supplementary data Cowx, I.G., O’Grady, K.T., O’Connor, W.C.K., Andrew, T.E., 1998. The effects of
siltation on Atlantic salmon, Salmo salar L., embryos in the River Bush. Fisheries
Manage. Ecol. 5, 393–401.
Supplementary data associated with this article can be found, in Dibble, K.L., Yackulic, C.B., Kennedy, T.A., Budy, P., 2015. Flow management and fish
the online version, at http://dx.doi.org/10.1016/j.jhydrol.2017.04. density regulate salmonid recruitment and adult size in tailwaters across
023. western North America. Ecol. Appl. 25, 2168–2179.
Directive 2009/28/EC of the European Parliament and of the Council of 23 April
2009 on the promotion of the use of energy from renewable sources. Brussels.
References Elliott, J.M., Hurley, M.A., Elliott, J.A., 1997. Variable effects of droughts on the
density of a sea-trout population over 30 years. J. Appl. Ecol. 34, 1229–1238.
Ernst, W.H.O., 1990. Ecophysiology of plants in waterlogged and flooded
Álvarez, L.V., Schmeeckle, M.W., 2013. Erosion of river sandbars by diurnal stage
environments. Aquat. Bot. 38, 73–90.
fluctuations in the Colorado River in the Marble and Grand Canyons: full-scale
European Greenpower Marketing. 2006. Hervorragende Aussichten für
laboratory experiments. River Res. Appl. 29, 839–854.
Wasserkraft: <www.greenpowermarkets.eu/>.
Amlin, N.M., Rood, S.B., 2002. Comparative tolerances of riparian willows and
Fitzhugh, T.W., Vogel, R.M., 2011. The impact of dams on flood flows in the United
cottonwoods to water-table decline. Wetlands 22, 338–346.
States. River Res. Appl. 27, 1192–1215.
Archer, D., Newson, M., 2002. The use of indices of flow variability in assessing the
Flodmark, L.E.W., Vollestad, L.A., Forseth, T., 2004. Performance of juvenile brown trout
hydrological and instream habitat impacts of upland afforestation and drainage.
exposed to fluctuating water level and temperature. J. Fish Biol. 65, 460–470.
J. Hydrol. 268, 244–258.
Freeman, M.C., Bowen, Z.H., Bovee, K.D., Irwin, E.R., 2001. Flow and habitat effects
Armstrong, W., Brändle, R., Jackson, M.B., 1994. Mechanisms of flood tolerance in
on juvenile fish abundance in natural and altered flow regimes. Ecol. Appl. 11,
plants. Acta Bot. Neerl. 43, 307–358.
179–190.
Armstrong, J.D., Kemp, P.S., Kennedy, G.J.A., Ladle, M., Milner, N.J., 2003. Habitat
Friedman, J.M., Auble, G.T., 1999. Mortality of riparian box elder from sediment
requirements of Atlantic salmon and brown trout in rivers and streams. Fish.
mobilization and extended inundation. Regul. Rivers- Res. Manage. 15, 463–476.
Res. 62, 143–170.
Gao, Y., Vogel, R.M., Kroll, C.N., Poff, N.L., Olden, J.D., 2009. Development of
Bailey-Serres, J., Voesenek, L.A.C.J., 2008. Flooding stress: acclimations and genetic
representative indicators of hydrologic alteration. J. Hydrol. 374, 136–147.
diversity. Ann. Rev. Plant Biol. 59, 313–339.
Garcia de Jalón, D., Montes, C., Barcelo, E., Casado, C., Menes, F., 1998. Effects of
Baker, D.B., Richards, R.P., Loftus, T.T., Kramer, J.W., 2004. A new flashiness index:
hydroelectric scheme on fluvial ecosystems within the Spanish Pyrenees. Regul.
characteristics and applications to midwestern rivers and streams. J. Am. Water
Rivers-Res. Manage. 2, 479–491.
Resour. Assoc. 40, 503–522.
Gasith, A., Resh, V.H., 1999. Streams in Mediterranean climate regions: abiotic
Baldwin, A.H., Egnotovich, M.S., Clarke, E., 2001. Hydrologic change and vegetation
influences and biotic responses to predictable seasonal events. Ann. Rev. Ecol.
of tidal freshwater marshes: field, greenhouse, and seed-bank experiments.
Syst. 30, 51–81.
Wetlands 21, 519–531.
Goodson, J.M., Gurnell, A.M., Angold, P.G., Morrissey, I.P., 2003. Evidence for
Bejarano, M.D., Marchamalo, M., de Jalón, D.G., del Tánago, M.G., 2010. Flow regime
hydrochory and the deposition of viable seeds within winter flow-deposited
patterns and their controlling factors in the Ebro basin (Spain). J. Hydrol. 385,
sediments: the River Dove, Derbyshire. UK. River Res. Appl. 19, 317–334.
323–335.
Gorla, L., Signarbieux, C., Turberg, P., Buttler, A., Perona, P., 2015. Transient response Meile, T., Boillat, J.L., Schleiss, A.J., 2011. Hydropeaking indicators for
of Salix cuttings to changing water level regimes. Water Resour. Res. 51, 1758– characterization of the Upper-Rhone River in Switzerland. Aqua Sci. 73, 171–
1774. 182.
Gostner, W., Lucarelli, C., Theiner, D., Kager, A., Premstaller, G., Schleiss, A., 2011. A Merritt, D.M., Nilsson, C., Jansson, R., 2010. Consequences of propagule dispersal
holistic approach to reduce negative impacts of hydropeaking. In: Schleiss, A.J., and river fragmentation for riparian plant community diversity and turnover.
Boes, R.N. (Eds.), Dams and Reservoirs Under Changing Challenges. CRC Press, Ecol. Monogr. 80, 609–626.
Boca Raton, pp. 857–866. Miller, S.W., Judson, S., 2014. Responses of macroinvertebrate drift, benthic
Grumbine, R.E., Xu, J., 2011. Mekong hydropower development. Science 332, 178– assemblages, and trout foraging to hydropeaking. Can. J. Fish. Aquat. Sci. 71,
179. 675–687.
Haas, N.A., O’Connor, B.L., Hayse, J.W., Bevelhimer, M.S., Endreny, T.A., 2014. Milner, N.J., Elliott, J.M., Armstrong, J.D., Gardiner, R., Welton, J.S., Ladle, M., 2003.
Analysis of daily peaking and run-of-river operations with flow variability The natural control of salmon and trout populations in streams. Fish. Res. 62,
metrics, considering subdaily to seasonal time scales. J. Am. Water Resour. Ass. 111–125.
50, 1622–1640. Mjelde, M., Hellsten, S., Ecke, F., 2013. A water level drawdown index for aquatic
Halleraker, J.H., Saltveit, S.J., Harby, A., Arnekleiv, J.V., Fjeldstad, H.P., Kohler, B., macrophytes in Nordic lakes. Hydrobiologia 704, 141–151.
2003. Factors influencing stranding of wild juvenile brown trout (Salmo trutta) Moliere, D.R., Lowry, J.B., Humphrey, C.L., 2009. Classifying the flow regime of data-
during rapid and frequent flow decreases in an artificial stream. River Res. Appl. limited streams in the wet-dry tropical region of Australia. J. Hydrol. 367, 1–13.
19, 589–603. Moog, O., 1993. Quantification of daily peak hydropower effects on aquatic fauna and
Harris, N.M., Gurnell, A.M., Hannah, D.M., Petts, G.E., 2000. Classification of river management to minimize environmental impacts. Regul. Rivers-Res. Manage. 8,
regimes: a context for hydroecology. Hydrol. Proc. 14, 2831–2848. 5–14.
Hauer, C., Unfer, G., Holzapfel, P., Haimann, M., Habersack, H., 2014. Impact of Nagrodski, A., Raby, G.D., Hasler, C.T., Taylor, M.K., Cooke, S.J., 2012. Fish stranding
channel bar form and grain size variability on estimated stranding risk of in freshwater systems: sources, consequences, and mitigation. J. Environ.
juvenile brown trout during hydropeaking. Earth Surf. Processes Landforms 39, Manage. 103, 133–141.
1622–1641. Nilsson, C., Ekblad, A., Gardfjell, M., Carlberg, B., 1991. Long-term effects of river
Havens, K.E., Sharfstein, B., Brady, M.A., East, T.L., Harwell, M.C., Maki, R.P., Rodusky, regulation on river margin vegetation. J. Appl. Ecol. 28, 963–987.
A.J., 2004. Recovery of submerged plants from high water stress in a large Olden, J.D., Poff, N.L., 2003. Redundancy and the choice of hydrologic indices for
subtropical lake in Florida, USA. Aquat. Bot. 78, 67–82. characterizing streamflow regimes. River Res. Appl. 19, 101–121.
Heggenes, J., Alfredsen, K., Adeva Bustos, A., Huusko, A., 2016. Be cool: hydro- Pettit, N.E., Naiman, R.J., 2006. Flood-deposited wood creates regeneration niches for
physical changes and fish responses in winter in hydropower-regulated riparian vegetation on a semi-arid South African river. J. Veg. Sci. 17, 615–624.
northern streams. Publication series from the University College of Southeast Poff, N.L., Allan, J.D., Bain, M.B., Karr, J.R., Prestegaard, K.L., Richter, B.D., Sparks, R.E.,
Norway no 1, Kongsberg. Stromberg, J.C., 1997. The natural flow regime: a paradigm for river
Holzapfel, P., Leitner, P., Habersack, H., Graf, W., Hauer, C., 2016. Evaluation of conservation and restoration. Bioscience 47, 769–784.
hydropeaking impacts on the food web in alpine streams based on modelling of Poff, N.L., Richter, B.D., Arthington, A.H., Bunn, S.E., Naiman, R.J., Kendy, E., Acreman,
fish-and macroinvertebrate habitats. Sci. Total Environ. 575, 1489–1502. M., Apse, C., Bledsoe, B.P., Freeman, M., Henriksen, J., Jacobson, R.B., Kennen, J.,
Irvine, R.L., Oussoren, T., Baxter, J.S., Schmidt, D.C., 2009. The effects of flow Merritt, D.M., O’Keeffe, J., Olden, J.D., Rogers, K., Tharme, R.E., Warner, A., 2010.
reduction rates on fish stranding in British Columbia, Canada. River Res. Appl. The ecological limits of hydrologic alteration (ELOHA): a new framework for
25, 405–415. developing regional environmental flow standards. Freshw. Biol. 55, 147–170.
Irvine, R.L., Thorley, J.L., Westcott, R., Schmidt, D., DeRosa, D., 2015. Why do fish Porporato, A., Laio, F., Ridolfi, L., Rodríguez-Iturbe, I., 2001. Plants in water-
strand? An analysis of ten years of flow reduction monitoring data from the controlled ecosystems: active role in hydrologic processes and response to
Columbia and Kootenay rivers, Canada. River Res. Appl. 31, 1242–1250. water stress: III. Vegetation water stress. Adv. Water Resour. 24, 725–744.
Jager, H.I., Efroymson, R.A., Opperman, J.J., Kelly, M.R., 2015. Spatial design Puffer, M., Berg, O.K., Huusko, A., Vehanen, T., Forseth, T., Einum, S., 2015. Seasonal
principles for sustainable hydropower development in river basins. effects of hydropeaking on growth, energetics and movement of juvenile
Renewable Sustainable Energy Rev. 45, 808–816. Atlantic salmon (Salmo salar). River Res. Appl. 31, 1101–1108.
Jensen, A.J., Johnsen, B.O., 1999. The functional relationship between peak spring Räsänen, T.A., Koponen, J., Lauri, H., Kummu, M., 2012. Downstream hydrological
floods and survival and growth of juvenile Atlantic salmon (Salmo salar) and impacts of hydropower development in the Upper Mekong Basin. Water Resour.
brown trout (Salmo trutta). Funct. Ecol. 13, 778–785. Manage 26, 3495–3513.
Kelly, B., Smokorowski, K.E., Power, M., 2017. Impact of river regulation and Richter, B., Baumgartner, J., Powell, J., Braun, D., 1996. A method for assessing
hydropeaking on the growth, condition and field metabolism of brook trout hydrologic alteration within ecosystems. Conserv. Biol. 10, 1163–1174.
(Salvelinus fontinalis). Ecol. Freshwater Fish. http://dx.doi.org/ Richter, B.D., Baumgartner, J.V., Wigington, R., Braun, D.P., 1997. How much water
10.1111/eff.12310. does a river need? Freshw. Biol. 37, 231–249.
Kennard, M.J., Pusey, B.J., Olden, J.D., MacKay, S.J., Stein, J.L., Marsh, N., 2010. Robertson, M.J., Pennell, C.J., Scruton, D.A., Robertson, G.J., Brown, J.A., 2004. Effect
Classification of natural flow regimes in Australia to support environmental of increased flow on the behaviour of Atlantic salmon parr in winter. J. Fish Biol.
flow management. Freshwater Biol. 55, 171–193. 65, 1070–1079.
Kennedy, T.A., Muehlbauer, J.D., Yackulic, C.B., Lytle, D.A., Miller, S.W., Dibble, K.L., Rocaspana, R., Aparicio, E., Vinyoles, D., Palau, A., 2016. Effects of pulsed discharges
Koertenhoeven, E.W., Metcalfe, A., Baxter, C.V., 2016. Flow management for from a hydropower station on summer diel feeding activity and diet of brown
hydropower extirpates aquatic insects, undermining river food webs. trout (Salmo trutta Linnaeus, 1758) in an Iberian stream. J. Appl. Ichthyol. 32,
Bioscience 66, 561–575. 190–197.
Kirk, J.T., 1994. Light and Photosynthesis in Aquatic Ecosystems. Cambridge Roni, P., Hanson, K., Beechie, T., 2008. Global review of the physical and biological
University Press, Cambridge. effectiveness of stream habitat rehabilitation techniques. N. Am. J. Fish. Manage.
Korman, J., Campana, S.E., 2009. Effects of hydropeaking on nearshore habitat use 28, 856–890.
and growth of age-0 rainbow trout in a large regulated river. Trans. Am. Fish. Saltveit, S.J., Halleraker, J.H., Arnekleiv, J.V., Harby, A., 2001. Field experiments on
Soc. 138, 76–87. stranding in juvenile Atlantic salmon (Salmo salar) and brown trout (Salmo
Kuenzer, C., Campbell, I., Roch, M., Leinenkugel, P., Tuan, V.Q., Dech, S., 2013. trutta) during rapid flow decreases caused by hydropeaking. Regul. Rivers-Res.
Understanding the impact of hydropower developments in the context of Manage. 17, 609–622.
upstream–downstream relations in the Mekong river basin. Sustainable Sci. 8, Sarneel, J.M., Janssen, R.H., Rip, W.J., Bender, I., Bakker, E.S., 2014. Windows of
565–584. opportunity for germination of riparian species after restoring water level
Lees, A.C., Peres, C.A., Fearnside, P.M., Schneider, M., Zuanon, J.A., 2016. fluctuations: a field experiment with controlled seed banks. J. Appl. Ecol. 51,
Hydropower and the future of Amazonian biodiversity. Biodivers. Conserv. 25, 1006–1014.
451–466. Sauterleute, J.F., Charmasson, J., 2014. A computational tool for the characterisation
Leitner, P., Hauer, C., Graf, W., 2017. Habitat use and tolerance levels of of rapid fluctuations in flow and stage in rivers caused by hydropeaking.
macroinvertebrates concerning hydraulic stress in hydropeaking rivers: a case Environ. Modell. Softw. 55, 266–278.
study at the Ziller River in Austria. Sci. Total Environ. 575, 112–118. Schuster, W.D., Zhang, Y., Roy, A.H., Daniel, F.B., Troyer, M., 2008. Characterizing
Levasseur, M., Bergeron, N.E., Lapointe, M.F., Bérubé, F., 2006. Effects of silt and very storm hydrograph rise and fall dynamics with stream gage data. J. Am. Water
fine sand dynamics in Atlantic salmon (Salmo salar) redds on embryo hatching Resour. Assoc. 44, 1431–1440.
success. Can. J. Fish. Aquat. Sci. 63, 1450–1459. Scruton, D.A., Ollerhead, L.M.N., Clarke, K.D., Pennell, C., Alfredsen, K., Harby, A.,
Lind, L., Nilsson, C., Polvi, L.E., Weber, C., 2014. The role of ice dynamics in shaping Kelley, D., 2003. The behavioural response of juvenile Atlantic salmon (Salmo
vegetation in flowing waters. Biol. Rev. 89, 791–804. salar) and brook trout (Salvelinus fontinalis) to experimental hydropeaking on a
Lundquist, J.D., Cayan, D.R., 2002. Seasonal and spatial patterns in diurnal cycles in Newfoundland (Canada) river. River Res. Appl. 19, 577–587.
streamflow in the western United States. J. Hydrometeorology 3, 591–603. Scruton, D.A., Pennell, C.J., Robertson, M.J., Ollerhead, L.M.N., Clarke, K.D., Alfredsen,
Lytle, D.H., Poff, N.L., 2004. Adaptation to natural flow regimes. Trends Ecol. Evol. 19, K., Harby, A., McKinley, R.S., 2005. Seasonal response of juvenile Atlantic salmon
94–100. to experimental hydropeaking power generation in Newfoundland, Canada. N.
Madsen, J.D., Chambers, P.A., James, W.F., Koch, E.W., Westlake, D.F., 2001. The Am. J. Fish. Manage. 25, 964–974.
interaction between water movement, sediment dynamics and submersed Scruton, D.A., Pennell, C., Ollerhead, L.M.N., Alfredsen, K., Stickler, M., Harby, A.,
macrophytes. Hydrobiologia 444, 71–84. Robertson, M., Clarke, K.D., LeDrew, L.J., 2008. A synopsis of ‘hydropeaking’
McManamay, R.A., Orth, D.J., Dolloff, C.A., Frimpong, E.A., 2012. A regional studies on the response of juvenile Atlantic salmon to experimental flow
classification of unregulated stream flows: spatial resolution and hierarchical alteration. Hydrobiologia 609, 263–275.
frameworks. River Res. Appl. 28, 1019–1033. Smith, C., Reay, P., 1991. Cannibalism in teleost fish. Rev. Fish Biol. Fish. 1, 41–64.
Stella, J.C., Battles, J.J., McBride, J.R., Orr, B.K., 2010. Riparian seedling mortality from Vehanen, T., Jurvelius, J., Lahti, M., 2005. Habitat utilisation by fish community in a
simulated water table recession, and the design of sustainable flow regimes on short-term regulated river reservoir. Hydrobiologia 545, 257–270.
regulated rivers. Restor. Ecol. 18, 284–294. Vollset, K.W., Skoglund, H., Wiers, T., Barlaup, B.T., 2016. Effects of hydropeaking on
Taylor, M.K., Cooke, S.J., 2012. Meta-analyses of the effects of river flow on fish the spawning behaviour of Atlantic salmon Salmo salar and brown trout Salmo
movement and activity. Environ. Rev. 20, 211–219. trutta. J. Fish Biol. 88, 2236–2250.
Taylor, M.K., Hasler, C.T., Hinch, S.G., Lewis, B., Schmidt, D.C., Cooke, S.J., 2014. Reach- Ward, J.V., 1989. The four-dimensional nature of lotic ecosystems. J. N. Am. Benthol.
scale movements of bull trout (Salvelinus confluentus) relative to hydropeaking Soc. 8, 2–8.
operations in the Columbia River, Canada. Ecohydrology 7, 1079–1086. Warren, M., Dunbar, M.J., Smith, C., 2015. River flow as a determinant of salmonid
Timusk, E.R., Smokorowski, K.E., Jones, N.E., 2016. An experimental test of sub- distribution and abundance: a review. Environ. Biol. Fish. 98, 1695–1717.
hourly changes in macroinvertebrate drift density associated with White, M.A., Schmidt, J.C., Topping, D.J., 2005. Application of wavelet analysis for
hydropeaking in a regulated river. J. Freshwater Ecol. 31, 555–570. monitoring the hydrologic effects of dam operation: Glen Canyon Dam and the
Tuhtan, J.A., Noack, M., Wieprecht, S., 2012. Estimating stranding risk due to Colorado River at Lees Ferry, Arizona. River Res. Appl. 21, 551–565.
hydropeaking for juvenile European grayling considering river morphology. Wolter, C., Sukhodolov, A., 2008. Random displacement versus habitat choice of fish
KSCE J. Civ. Eng. 16, 197–206. larvae in rivers. River Res. Appl. 24, 661–672.
Van Looy, K., Jochems, H., Vanacker, S., Lommelen, E., 2007. Hydropeaking impact Wootton, R.J., 1998. Ecology of Teleost Fishes. Kluwer Academic Publishers,
on a riparian ground beetle community. River Res. Appl. 23, 223–233. Dordrecht.
Vanzo, D., Zolezzi, G., Siviglia, A., 2015. Eco-hydraulic modelling of the Zimmerman, J.K., Letcher, B.H., Nislow, K.H., Lutz, K.A., Magilligan, F.J., 2010.
interactions between hydropeaking and river morphology. Ecohydrology 9, Determining the effects of dams on subdaily variation in river flows at a whole-
421–437. basin scale. River Res. Appl. 26, 1246–1260.